Creating Vanilla WoW Dungeons - Learn More in the WoW Diary Kickstarter
We have another article on Classic WoW for you by John Staats, the creator of the
WoW Diary Kickstarter
. In this article, John describes the creation of all vanilla dungeons he worked on--an exclusive article for Wowhead which won't be found in the
book. Check out our previous exclusive articles as well on the
creation of Scholomance
Click here to support
The WoW Diary
Kickstarter! Less than a week left to pledge.
About the Author and Kickstarter
A veteran of WoW's development, John Staats in vanilla designed and built Ahn’Qiraj Temple, Blackfathom Deeps, Blackwing Lair, Blackrock Mountain, Blackrock Depths, Booty Bay, Karazhan (w/Aaron Keller), Loch Modan Dam, Lower Blackrock Spire, Molten Core, Razorfen Downs, Razorfen Kraul, Scholomance, The Slag Pit, Upper Blackrock Spire, The Wailing Caverns, and Warsong Gulch (w/Matt Milizia).
When John first joined Blizzard Entertainment, he was unfamiliar with the games industry and took copious notes on everything he learned and even interviewed his teammates. These detailed notes serve as the basis for this new book! This book includes not just essays on game development and fun anecdotes about the WoW team, but over 130 annotated images printed on high-quality paper stock.
The book will be available to
Kickstarter backers first
, with pledges starting at $5, electronic copies at $30, and physical copy for $40. There are also deluxe options available--$150 nets you supplemental essays on WoW's development. All Kickstarter products are expected to ship by December 2018.
If you're an avid WoW player with fond memories of the early days, or want to learn more about the games industry in the hopes of joining it one day, this book definitely seems for you!
The WoW Diary
has recently crossed the first stretch goal at $300k! Check out the other goals that are within reach:
This stretch goal will fund two versions of
The WoW Diary
. The first, will be the standard Amazon Edition that has all the high-end specs already listed. Then, I'll print a Golden Edition that will be for only and all of my Kickstarter First Edition backers.
Upgrade 1: Unlocks at $300,000 The end sheet (pictured below) will be gold that reads “Original Kickstarter Backer" followed the whenitsready logo. The mock-up (on the right) shows the subtle printing, and the magnification (on the left) is darkened and enlarged for readability.
Upgrade 2: Unlocks at $350,000 - Gold foil title lettering on spine and front cover differentiates Kickstarter editions from non-KS prints.
Upgrade 3: Unlocks at $400,000 - Spot varnishing makes all the pictures shiny and opens opportunity for shiny "ghost shapes" on an otherwise blank page. Very few books have this because it costs a lot, but if we hit this mark, I'll add it for free.
Pushing Polys - My Vanilla WoW Dungeons
For this week's article, John Staats includes notes about all the dungeons, as well as some other zones, he worked on for classic WoW. While these anecdotes are interesting, they're not all specifically about the pre-launch phase of development, so they are left out of the book.
This is a monster-sized essay and we are greatly indebted to John for sharing so many of his observations with the Wowhead community! Here are some of the highlights:
John ranks Ahn'Qiraj as his worst instance. There were issues due to a shortage of texture artists, unfamiliarity with 3D Studio Max, and lack of experience designing interior environments without skyboxes or nature to break up the space. He also envisioned the instance, from above, resembling a scarab--but this caused many design problems and after all that work, raiders were unaware of the visual reference.
Blackfathom Deeps was purposely designed to feel claustrophobic through optical illusions.
John feels an odd sense of kinship with Franclorn Forgewright, the Warcraft architect credited in-game for building many of the structures John was responsible for such as the Loch Modan Dam and Shadowforge City.
Alex Afrasiabi ran into unique difficulties when running over the lava on the way back to Molten Core. He played a Tauren, and the larger collusion box for that race caused him to fall into the lava frequently when running on the chain. There was no way out of the lava during testing, so he'd have to release back to the grayeyard, further damaging his armor. Adding a way out of the lava was one of the last ninja-hotfixes John made before WoW went live.
Blackrock Depths is John's favorite dungeon, as the layout felt believable as a real functioning city. While he's not happy with The Lyceum, a room added late in the dungeon's development, the game designers loved the room and created an innovative encounter.
Booty Bay was the first test of WoW's engine, demonstrating to the team they could build complicated cities without having to worry about rendering issues or framerate drops.
The Loch Modan Dam was modeled after Hoover Dam. It was a very-late insertion and ruffled some feathers as Loch Modan was considered "complete" at the time. However, the team soon appreciated the extra work and dwarves touches that went into the new dam.
Molten Core's layout was designed in a few days and it was built "off the menu" as Team 2 had no spare resources to work on the raid.
Razorfen Downs and Razorfen Kraul are shorter than the other vanilla dungeons because Blizzard was unable to expand the size of the dungeons due to the limitations of the canopies.
Jeff Kaplan was most impressed by the Beast in UBRS, which in reality was one of the fastest parts of the dungeon to put together.
The Wailing Caverns internal playlets was one of the few dungeon tests employees conducted. It was a fiasco where everyone's armor turned red, nobody knew what that meant, and they didn't make it to even a subboss.
Ahn’Qiraj Temple (raid)
The WoW team originally built architecture the same way first-person shooter devs did, but nearly everything we mimicked in their production pipeline was unusable for MMOs. I dedicated a chapter in my book (entitled “Nine Months Down the Tubes”) to these painful lessons. We began building geometry with a common development tool (that I’d never used before), called 3D Studio Max. I had no idea how many polygons I could use to build my areas because we couldn’t yet test my geometry in-game; nor did I understand how to best sculpt my 3D wireframe mesh to suit texture application. The 3D level designers began once again, taking stabs in the dark within a near-vacuum of feedback. Dead ends, mistakes, and lessons were to remain in my horoscope.
Ahn’Qiraj was easily my worst dungeon, which is not coincidental, in that it was also the first dungeon I’d built using 3D Studio Max. The resulting geometric mesh was very difficult to work with, and I had built the raid, and all the hive tunnels, the same way, leaving the team with the choice of starting the dungeon over, or using them as is.
Even after we abandoned the FPS approach to building geometry, applying textures to dungeons had always been a headache for the WoW dev team. Finding texture artists who specialized in buildings and environments was hard because it required a different skillset. Back in 2000, when studios were only beginning to abandon 2D games, experienced 3D texture artists were hard to find. Another issue with dungeons was that there were two people working on the same asset (instead of one), and the 3D level designer often builds the entire thing months before the texturer has time to begin their job. This type of disconnect is sometimes a knot in the production pipeline. Other assets (characters, monsters, props) don’t suffer from the same complication: Artist who model a prop (or doodad), paints and applies its texture immediately afterward. Even Gary Platner, who painted WoW’s terrain textures, could create demo areas and rapidly see his results.
Unlike the exterior terrain or props, architectural texturing defines the entire environment. There are no trees or skybox to give the eye a visual rest. To evaluate a room, the floor, wall, trim, and ceiling textures work together, they all must be painted and applied before they can be judged with a critical eye. Some colors clashed, were too busy, and were repetitive. Identifying these weaknesses was foreign process to nearly everyone on the Blizzard art staff. At the time, none of Blizzard’s artists had experience in working with assets that used many textures. They didn’t know how to coordinate them together. Everyone on staff found it too frustrating to make the transition. That’s why Blizzard needed outside help—which was hard to find.
When I built Ahn’Qiraj, I used a simple gray texture on the geometry. As I was new to 3D Studio Max, I’d made the mistake of sculpting all the details into the geometry. The waving, irregular 3D shapes of the tunnels made work difficult for the texture artists (when we’d hired them, almost a year later). In order to compete with my irregular surfaces, they painted their textures using heavy, saturated colors. Such textures prevented subtle colored lights from having any effect, so I couldn’t distinguish different sections of the dungeon using lighting (or even props). This led to an overall problem of the hive tunnels all looking alike, which made them disorienting. Players getting lost was one of our worst fears too. It didn’t help that the entire texture count was only a dozen or so, far fewer than most dungeons.
As bad as the hive tunnels were, the temple had other problems. I’d stolen an architectural feature from a Cyan game, called Riven, which had a gold dome with a slice going through its structure. The slice looked so cool on the outside, Cyan featured it in their promotional images. My intent was to cut two grooves into an oval structure, forming form a scarab from a bird’s-eye view. This jived with Chris Metzen’s direction that the temple should only hint at being Egyptian. But grooves cutting through the building meant I couldn’t connect the floorplan together, so most of the spaces inside were awkward and useless dead ends. I was bending over backwards to pull off the scarab/groove concept, and as it turned out, it wasn’t even something players saw. Dead ends like these are symptomatic of assets built early in a project. In every sense, I’d figuratively “painted myself into a corner.”
It’s no exaggeration to say that Ahn’Qiraj was unpopular with the dev team when it was placed in the world, spawned, and scripted a few years later. I’d revisited and expanded this dungeon so many times, trying to improve its flawed concepts and geometry, I’d become grouchy about it. I was tired of making ineffectual lighting and geometry adjustments; the texture artists were weary of my asks for additional textures, and eventually the game designers were stuck with a long, narrow play spaces, mostly unsuitable for 40-player raids. They even allowed players to mount inside the tunnels to alleviate the travel time.
Blackfathom Deeps (dungeon)
Blackfathom was a rarely-visited dungeon. I was a happy with the entrance, I thought the spiraling ramps and swim-through were a unique way to go underground. My goal was to make a cave that felt different than the Wailing Caverns, I wanted something claustrophobic, and I achieved this not by making the ceiling lower, but by widening the passages. The optical illusion fooled some developers in thinking the ceiling was so low that it would interfere with the player’s camera. Another unsettling idea was the ceiling and walls didn’t quite touch each other. It was an effect I achieved by deepening the corners where they’d normally touch. I think the overall effect was that this dungeon felt more oppressive than it needed to be…but it did contrast with the game’s other caves. At least it contrasted with the white temple architecture, making it a welcome change.
I am willing to bet lot of level designers would agree that working with classical architecture isn’t easy. The familiar, symmetrical style of Greek temples doesn’t make for exciting play spaces, although breaking them into ruins availed me of some interesting artistic opportunities. The most memorable scenes were the waterfalls, and gorgeous wisp props that gave this dungeon its flavor. Blackfathom was also the site of The Great Transparent Water Controversy (see the chapter “Stepping on Toes,” on page 254 of
The WoW Diary
). In my opinion, the use of water is what made this dungeon successful.
Blackwing Lair (raid)
This was the last Blackrock dungeon I’d built, so I was exhausted of ideas for dark iron dwarf architecture (I’d been using the same texture set over a year in building BRM, UBRS, LBRS, and BRD). In fact, I was started near the end of the dev cycle, so I was burnt out in general. With no foreseeable release date of the raid, I build a few rooms and tabled it for later. After Vanilla WoW shipped, I expanded the size of Blackwing’s Lair and artists became available to create custom-made props (a rare luxury for dungeons) for Flamegor’s Laboratory.
I was able to sit in on encounter team meetings and see how they designed raids and scripted boss fights. I liked how they integrated my geometry into the fight. This insight was helpful because dungeons were almost always built long before the encounter designers designed fights. I noticed now they used decorative geometry as combat elements. For instance, Razorgore’s battle employed a kiting technique between two raised platforms (mirroring rooms speeds up the building process), and they spawned monsters from alcoves that I’d worried might interfere with combat. Instead, the scripters said, “We love it! The room is perfect! Don’t touch anything!” In other rooms, players used the geometry to block line-of-sight damage, such as Firemaw, which admittedly, felt broken. The scripters enjoyed the fact that random room-layouts and environmental elements forced them to be creative with fight ideas. While a lack of planning helped to keep boss fights fresh, some rooms were agnostic to combat. Such examples were the suppression room (which “gave raid rogues something fun to do”) and the crazy DPS race of Vaelastrasz.
Thematically, I was caught between lore and gameplay: Chris Metzen’s vision for the space called for the tallest structure in the game (so I wanted to emphasize the feel the players were always ascending). Yet the game designers were calling for flat play spaces, far away from stairs (where line of sight breaks down and players could take advantage of slow monster pathing). Artistically, I would rate the overall architecture as ho-hum, although I was happy that the boss room overlooked the exterior terrain. I got some pushback from the balcony concept. People weren’t convinced that connecting people to the outside was important enough to deal with issues of letting players see a “façade” of the exterior world below. It created work for other people, but I felt it was necessary to pull the illusion of climbing the world’s tallest spire. Besides, letting players outside was a refreshing way to punctuate an experience as tough as a raid dungeon, and I felt we hadn’t let players in dungeons see “outside” enough (Shadowfang was the only other instance that I can remember).
Blackrock Mountain (world location)
Blackrock Mountain was easily my favorite creation and I covered its origins in
The WoW Diary
, pages 175-176. BRM was special for a few reasons. I built it right before we announced the game and I was proud that it featured prominently in our gameplay trailer. It made me feel like I really contributed something to the project, like the scope of it had set a new bar for what our rooms could be. It was far more ambitious than the meticulous FPS levels I’d designed before coming to Blizzard.
My personal connection to the area got a little weird with the tomb of Franclorn Forgewright, the architect who’d supposedly designed the place. I got the idea from Chris Metzen, who thought it would be cool that this NPC would be interred in such an honorary location. Because Forgewright was also credited to constructing the Loch Modan dam and Shadowforge City (which were other things I’d built), coworkers joked that “I was buried there.” The parallels between me and the dead dwarf always elicited an unsettling apprehension that I was running past my own grave whenever I went to the Molten Core. The coincidence was unforgivingly corny, but the unease never waned.
What I didn’t mention in my book was this level was the site of the last ninja-fix I’d done in Vanilla WoW, and the very last art change before the game went live. Alex Afrasiabi, one of WoW quest designers and leader of the preeminent
guild, Fires of Heaven, was raid-testing the Molten Core. He wanted to evaluate the raid as much as he could before it became open to the public. Late one night he showed up in my office with a wild look in his eyes. He begged me to change the geometry to allow players, who’d fallen into the lava in the big central chamber, to reach safety on shore. As is, there was no way to get out of the lava. Apparently, Alex kept falling into the lava because there was an invisible collision bump on the chain-bridge leading to the Molten Core. After his raid wiped, Alex would run back (as a ghost), but the trip was so long he’d forget about the bump…causing him to fall into lava. As a ghost, he couldn’t exit without releasing directly to the graveyard, further damaging his armor. Alex was one of the few raiders in his guild to play a tauren (whose collision boxes were larger), so it was only he who kept repeating the blunder. Alex’s raids were near the tail end of WoW’s dev cycle, when he was sleep deprived and fatigued. Wiping to raid bosses were stressful affairs (his guild, Fires of Heaven, took raiding seriously), so the embarrassment of falling into the lava added to his frustration. As I made the change, we both laughed after he admitted he’d cursed my name whenever he fell. We both knew it was a fairly dangerous to make art changes, since they could somehow introduce other bugs: But it was nearing midnight, and we were feeling lackadaisical about due process. I obliged him without alerting the attention of the QA department and producers (who had explicitly warned us to not make art changes without their approval). This ninja was the very last art change before WoW went gold the next morning. Alex’s pain was not for naught.
Blackrock Depths (dungeon)
This is my very favorite dungeon and it features some of my most scenic panoramas. The nonlinear layout was refreshing, and the interconnectivity between the slave mines, the prisons, gladiator ring, and city thematically clicked together with very few phony transitions (such as a prison next to a throne room). Its rooms had a strong sense of purpose and place. I amused myself by making multiple golem workshops, the last of which I dubbed the “Pour Your Own Lava Golem Room.” Forcing players to “sneak” down giant chains to the dungeon made them feel like they were entering areas where they didn’t belong. It all fed the element of immersion. I’m still proud at how believable this dark iron dwarf city feels. Adding to the immersion were environmental game mechanics, like players unlocking prison cells, and inventing a second way to pass through a single pair of doors. Reflecting gameplay into the geometry is one of my favorite level design pursuits.
Flamelash’s ledge was one of my favorite play-spaces, and the road that led to the Molten Core provided a dramatic backdrop. It was the original final boss location, but the encounter designers worried shortcuts to it could cheapen the loot. To lengthen the trip to the final encounter, I added three fairly ordinary rooms. I was out of ideas, so I made them processional hallways (wholly incongruous to the natural lava-hewn cavern to which they’re connected). I’m not a fan of mirrored rooms; they are generally a lazy way to build, and they disorient players. But I was burnt out, so I tacked on a few generic rooms, the “cheapest” of which was The Lyceum. Oddly enough, the game designers loved the Lyceum’s geometry, and they devised a hellish encounter involving a series of patrols and adds, one that wiped all but the best dungeoneers.
As an interesting note, the game’s lead dungeon scripter (who came from team 1 near WoW’s dev cycle), Scott Mercer, admitted that the quest reward for the dungeon’s final quest way too powerful. In fact, it was likely the most powerful reward ever given in WoW: A ring that granted an additional two percent chance to critically hit. Scott realized the problem himself, only after the game launched. A two percent chance to crit was better than most talent ultimates, and we couldn’t reward players with anything better (aside from an unsustainable progression of +3%, +4%, +5%, etc.). Scott explained that rewards too good effectively kill the slot they fit into. He couldn’t believe how overpowered it was, “In
, we’d give players a bonus item and think nothing of it.” Luckily, Scott never had to nerf the reward for “the princess quest” outright, since we were introducing a critical strike rating (which compared an item’s level to opponent’s level), a system that would ensure item designers could always give better rings at higher levels.
Booty Bay (world location)
“Straight up, Monkey Island.” That was the only art direction I was given, that was only the art direction I needed. Monkey Island was a game I’d never played (I never admitted this at the time) but I knew about the popular pirate cove archetype. Art director Bill Petras worked closely with me on this, possibly because it was an effort undertaken early in the project, possibly because he was a fan of Monkey Island, possibly because he was so geeked-out to see pirates in our game. Booty Bay was an area everyone cared about, it was flavorful, fun, and the team was excited to see something big without having to wait for programming support like our dungeons and cities.
Booty Bay also taught us many things about our engine…how many polygons it could draw; how many textures and batches it could load and render; how many props it would take to make a convincing settlement. All of these numbers told us how detailed we could build the rest of our locations, so I worked in the spotlight. I’d resisted advice to build it like Stormwind, using solid walls of buildings to occlude geometry (occluding geometry reduces drag on graphics cards). Based on my experience with other 3D engines, I was unconvinced we needed to be so conservative, and I thought the charm of a shanty-town would be ruined by solid, view-blocking structures. Still there were rumbles of concern about our framerate, so Scott Hartin, the game’s engine programmer, and I decided to hide it inside a bowl of exterior terrain. Occluding Booty Bay behind hollowed-out hills prevented it from clogging up video cards of players in Stranglethorn. The bowl idea let me stack the city’s structures vertically, like the aggregate village of Sweethaven from the movie, Popeye. I liked the idea of a big jungle gym where players could jump from level to level and it was something the game hadn’t seen before.
Bill also liked the idea of an entrance that revealed a surprise behind a rather modest tunnel—which also reinforced the hole-in-the-wall concept of a pirate town. We rearranged buildings around the tunnel over and over, maximizing the effect of a grand reveal. There was much discussion on whether the entrance should be high or low to maximize the awe of the player’s first view. Which would be more dramatic—looking up to see the city or looking down? I grew so tired of the debate, I didn’t care who made the decision (or why), but somewhere along the line, the quorum formed that a high entrance would have the most immediate impression, since the player-camera predominantly pointed downward.
It takes a village to make one. I followed concept sketches of artists Carlo Arellano and Tom Jung, who provided ideas for the feel for town, including the ship-inn. Carlo even made a prop (his first, I think) of a shark hanging from a rope, a likely tribute to
. Our newest hire, a desperately-needed architectural texture artist, Jimmy Lo, painted the entire Booty Bay texture set a month after I finished the geometry. Fresh out of school, Jimmy jumped straight into the colorful Warcraft palette, and created a terrific cartoony style, as polished as any industry veteran.
I built the transport ship at the same time, since it used similar textures (again, reducing the load on video cards). I rebuild it more than once. The first version didn’t even fit between the docks. I reduced it from its original, gigantic, “Warcraft proportions.” Carlo drew a terrific concept sketch for a fantastic, over-the-top ship. My problem with his concept was the sails weren’t “correct.” (I still cringe as I write this, sixteen years later.) I had recently taken sailing classes in Newport Beach and wanted to apply my maritime knowledge. I wanted to build a ship whose sails looked functional, even to a discerning eye, one whose sail shapes would create the proper propulsions (Nerd alert!). I ultimately went with an nautically correct vessel; and to this day, I still don’t know if I made the right decision.
The great news about Booty Bay is that it gave us proof that we didn’t need sight-blockers to save our framerate. We moved forward with more confidence that our buildings weren’t going to melt our rendering engine. That confidence was important because it meant our we wouldn’t need to severely limit our city designs…and that we could build big, complicated architectural structures without worrying that they’d be thrown away. We stopped building zig-zag entrances that separated our interiors from our exterior play spaces (zig-zags between rooms were common occlusion methods in FPS titles of the day). The only remaining zig-zag thresholds were the human inn (another building which we guessed would be a high-traffic area) and the strange, run-around wall exiting Northshire, where visual separation from the Elwynn Forest might be needed (in case player-traffic made it necessary to convert the newbie zone into an instance).
Karazhan (w/Aaron Keller) (dungeon)
Although it shipped with The Burning Crusade, Karazhan was center stage a few times during Vanilla WoW. I wrote about this dungeon extensively in
The WoW Diary
, in the section called, “Dungeons: The Last Hurdle,” starting on page 260.
Loch Modan Dam (world location)
This was another of my “ninja jobs” near the end of the project. The original dam had been built and placed in Loch Modan. That it towered over the waterline always bugged me. Some liked the “epic” height, but it looked like a bridge, and the structure blocked not only the view of the Wetlands, but the sky above the Wetlands!
I decided to create an optional dam, one that looked more functional. I unashamedly modeled mine after Hoover Dam. The only problem with this approach was its appearance from the Wetlands: It was just a flat, colorless, uninterrupted wall of cement. It looked bland and out of step with our game’s art direction. Since the dam was supposedly built by dwarves, I stuck giant dwarven faces (I stolen the face from the dwarf statues in Blackrock Mountain). It occurred to me that these faces could squirt water, like a fountain, which converted these decorative elements into believable spillways. The new pitch appealed to Bill Petras, and the switch was approved. But since it was late in the dev cycle, when everyone was tired and grouchy, gripes erupted from the exterior level design department. Some weren’t happy about reworking an area that was previously regarded to be “finished.” Changes to zones that were “locked down” was unsettling to some of my fatigued teammates. After a couple days of walking on eggshells, even they came around to agree the new one was worth the effort.
Lower Blackrock Spire (dungeon)
I am very happy with this “Swiss cheese dungeon,” a term Jeff Kaplan dubbed for large rooms which were entered repeatedly at different angles. I’d used the Swiss cheese technique for the big side wings of The Wailing Caverns, where players wound around large central chambers. It provided an immediately cool view, and one that changed as players progressed throughout the space. Another reason why this dungeon works so well for me was its ceiling. A ceiling that disappears into blackness, or is created from an undefined shape, is unsettling somehow. An artist I admired, H.R. Giger, once said some of his architectural work drew from childhood memories of big houses, where shafts, disappearing into darkness, conjured imaginative places. Perhaps it’s unsettling because our minds naturally want to complete things, or maybe players are paranoid enough to anticipate things dropping down upon them. Maybe the irregular shape of the main chamber suggests the whole thing could collapse at any moment. Either way, I liked that it reinforced a dangerous feel of exploring things off the beaten path.
The dichotomy of orcs infesting dwarven architecture somewhat paralleled the look of the Moria, from the recent film,
Fellowship of the Ring
. My only planning involved a quick sketch on graph paper, a practice I learned from making home-made adventures for Advanced Dungeons & Dragons. LBRS practically built itself. It was fun to “rubble out” architecture because ruined structures made for unusual play spaces, including peek-a-boo views into other rooms, drop-downs, and irregular slopes of rubble. This malleable approach was a nice change from the regular rhythm of rooms-and-hallways. Knocking out windows, from one room to another, provides perifery that parallaxed in the background, and it worked both ways. I especially like areas with a lived-in feel, one that has a sense of history to it, that used to serve a purpose to an abandoned city.
Molten Core (raid)
I covered the testing of The Molten Core in
The WoW Diary
, in the chapter, “World’s End,” pages 305-309. I expanded on the actual “production pipeline” of MC below:
Jeff Kaplan wanted a raid dungeon bigger than Onyxia’s Lair. Team 2 didn’t have the resources to make one before the game shipped. He solicited me to build one, and Scott Mercer (a scripter on team 1) to prove we could ninja something together real fast. I already had a license to build cool stuff “off the menu” and Scott (a die-hard
junkie) technically wasn’t on team 2, so Jeff figured we could build endgame material before WoW launched, not after. He just needed to prove it could be done (without alarming the producers). That’s usually the best way to ninja something into the build, do it, then getting permission from the producers. But to do it, Jeff needed free hands to execute it, which is where Scott and I came in. Jeff asking us to clandestinely build stuff was like offering food to the family dogs under the dinner table: There was no way either of us was going to turn it down. I built the basic layout of the dungeon in a few days, threw on some textures, and presented it as a proof-of-concept. We got the ball rolling so quickly, the producers shoehorned it onto the team’s official workload, one that gave me another couple weeks to do some polishing.
Razorfen Downs (dungeon)
Originally, there was only one Razorfen dungeon, but Chris Metzen suggested we spit it into two areas, and I eagerly supported the idea. I was keenly interested in building trenches instead of tunnels, and my visits to old American Civil War earthworks in Virginia fired my imagination on how these boars could build their living spaces.
Metzen’s concept of “a boar city beneath a giant thorn bush” was tricky to execute. To pull off the canopy, I went back and forth with our engine programmer, Scott Hartin, on how it could be done with our engine. Looking back, it’s a little weird to say, Scott was one of my closest collaborators on the project. He usually worked on weekends and made himself available for tweaking the limits of what his engine could do. He took time to explain how to building things, and sometimes hacked in solutions to our more ambitions visions. As an
enthusiast, Scott was sympathetic to the goal of making WoW epic. Having an engineer as an advocate was crucial for what the art and lore teams were trying to accomplish. Not all programmers would put in the extra work to do this, some could simply veto ambitious concepts. The game would have been very different without that kind of enthusiasm. Another helpful contributor to the Razorfen dungeons was Aaron Keller, a fellow 3D level designer. He taught me how to use loft objects in 3D Studio Max, a trick that helped me to achieve the solid thorn trunks. Without the canopy and an efficient way to make solid vines, this dungeon concept would likely have been scrapped.
This dungeon started from a terrific entrance concept sketch, by Carlo Arellano. Carlo gave me a lot of great ideas. He came up with the idea of hollowing out rooms within giant doors (from his concept sketch for Black Rock Depths). For Razorfen, he imagined an entrance that looked like a giant boar’s head. As a level designer, I’m usually resistant to the trope of mouth-entrances. They are a cliché in fantasy circles. In Gnomeregan, players walk into a mouth of a giant gnome. For Wailing Caverns, players walk into a moth of a skull-rock formation. I didn’t want our architecture’s first impression to become mouth-entrances of whatever enemies the players were soon to fight. Objections like these make me a pain in the ass to work with, but in this case, Carlo’s gaping boar maw was so cool, it changed my mind.
Razorfen Kraul (dungeon)
The crone section of the boars was meant to feel less marital than Razorfen Downs, but many of the artists weren’t happy with the similarity. I wanted to differentiate Razorfen Kraul by giving it gray dirt, something that looked sickly or dead, but repainting textures to a colorless palette wasn’t a popular idea with the art staff. I made the Kraul thorns skinnier and curved, as if they had withered, like a crone (the end boss), but this change wasn’t enough to help distinguish it from its sister dungeon, Razorfen Downs. Like the hive dungeons, the small texture set was a limiting factor. I suppose I could have built more cave sections, but that would have watered-down the thorn bush theme. We eventually settled for the repetitive feel.
Midway through development, it became a general worry that our players would go through content too quickly, so the game designers requested that our dungeons be made larger. This happened to The Wailing Caverns a number of times. All the Blackrock dungeons got longer. But two dungeons that didn’t get bigger were the two Razorfens because, thankfully, their canopies were a limiting factor.
Everything anyone could possibly want to know about the Scholomance can be found here.
The Slag Pit (world location)
My notes about the Slag Pit and the Searing Gorge structures in the “Art and Zones” section of
The WoW Diary
on page 119.
Upper Blackrock Spire (dungeon)
I’m not entirely happy with this dungeon’s architecture. It’s OK, but it’s not particularly interesting…which is a shame because it was our ultimate, pre-raid dungeon. It was the first of the Blackrock dungeons that I’d built, fairly early in the project, and its only lore was “dwarven architecture that it spiraled upwards.” Like all the dungeons, it was built before group combat had been tested. Game designers were worried nearby stairs would be awkward places to fight or where monster pathing might be exploited. I was caught between clashing concepts of elevation changes: Lore wanted it, design didn’t. I couldn’t make any big rooms overlooking one another because we were building under the early consensus that we needed to avoid letting players jump or fall to their deaths. Falling was a concern because we didn’t know if a break in internet connectivity would result in characters accidentally running off cliffs or over staircases (as it turned out, this wasn’t the case).
My favorite room was the rookery, and my lunchtime pal and quest designer, Pat Nagel, turned it into fertile ground for hijinks: By placing eggs everywhere, he enticed a certain Leroy Jenkins to achieve overnight internet fame. Over objections from some game designers, I created a drop-down (from the rookery) that lead outside the instance. It was an escape for any players overwhelmed by adds, and it was the few examples of multiple dungeon exits.
For a couple years (before it was spawned), the original final boss room was in Warchief Rend Blackhand’s location, next to the entrance of Blackwing’s Lair. Since the game designers said Lower Blackrock Spire and Blackrock Depths were so much more impressive, they thought UBRS needed a little grandeur…and I completely agreed. My solution was simple and efficient: I decided to simply connect UBRS to LBRS. It gave a vertiginous peek into the other dungeon for relatively little effort. I built the bridge over LBRS because we weren’t so obsessively worried anymore about disconnected players accidentally falling to their deaths. Connecting the dungeons together forced them into becoming the same instance. This made the game designers nervous, since we’d never done this before, but they couldn’t think of a way it could be exploited, so the link remained. Jeff Kaplan made one last request: He wanted a dark cubby hole, from where a monster called The Beast would emerge. It took five minutes to make but he was ecstatic at the result, “It’s so perfect! It exactly as dark as we want—and it’s the perfect size!” After Jeff left our office, all the level designers laughed and shook our heads. After months of designing and building geometry, applying textures, placing props, and lighting, the most acclaim we elicited from the team came from the simplest tweaks.
The Wailing Caverns (dungeon)
This dungeon. I devoted a whole section to it in
The WoW Diary
called, “The Growing Pains of The Wailing Caverns” on page 205.
A memory comes to mind of my first and last attempt to coerce the interior level design team to experience dungeon gameplay. During WoW’s alpha test, I was the only one in our department with the MMO-like patience to play through our many-hour instances, and I thought it important to familiarize everyone with grouped gameplay. I convinced my fellow dungeon artists to join me in an internal playtest.
Although there was no loot yet, I wanted to see how the Wailing Caverns played. It was The Shiny New Thing because the encounter designers had just finished scripting, debugging, and balancing, and spawning the area with monsters. After an hour of getting everyone onto the same build, on a stable internal server, into mid-level character, everyone’s patience was already thinning. No one wanted to listen to instructions or explanations about tanking (a mechanic unfamiliar to solo players in the alpha test), targeted pulls, aggro-management, or crowd-control. Many didn’t recognize their spells (20th level was higher than anyone had played in the alpha). My chat text wasn’t so much ignored as it was unseen—I made the wrong assumption that everyone paid attention to the chat interface. Furthermore, the newest feature, durability, had been implemented, using a coefficient that was so high that just a half hour of fighting rendered our equipment broken…and since it was a new feature, we didn’t know what
meant. We figured it was just a visual bug. The cheats for repairing armor were unknown (we didn’t realize there was such a thing as durability), so we experienced over an hour of utter frustration and failure. We didn’t even make it to a subboss.
The debacle ended with undisguised animosity toward party-based gameplay. The experience had turned the dungeon artists
playing dungeons. Despite the fact they were incredibly talented developers, the act of methodically picking apart monster groups, for hours on end, didn’t even sound fun to them. Only until the game was live, on servers, did anyone else in the dungeon department attempt to play through their own environments.
Warsong Gulch (w/Matt Milizia) (PvP battleground)
Everything I wanted to say about the Warsong Gulch can be found here.
Microdungeons (non-instanced, world dungeons)
Custom-made goldmines, caves, shipwrecks, and hive tunnels were, arguably, my most important contribution to the project. I talked about their evolution throughout
The WoW Diary
This article is like walking through Hannibal Lector’s garage, wherein Hester Mophet stores the rest of me. It contains various odds and ends, too valuable to throw away, too useless to belong anywhere else. I left most of these details out of
The WoW Diary
because the damned book was over 95,000 words already; and while these anecdotes may be interesting, they didn’t particularly contribute to the overall narrative of making the game. After all, WoW is more than a bunch of dungeons.
For a story that encompasses all aspects of game development, my book is on
for one last week.
Check it out!
获取 Wowhead 高级会员
[As little as
less than $1 a month
to enjoy an ad-free experience, unlock premium features, and support the site!]
[Sign In to Post a Comment]
其他 Fanbyte 站点
Hearthstone Top Decks
Final Fantasy XI
© 2021 Fanbyte